Experiments are being conducted to assess the different neuropharmacological and behavioral mechanisms underlying behavior controlled by drugs as discriminative stimuli in rats and monkeys and the ability of pharmacological or behavioral manipulations to modify such behavior. Currently, studies are focusing on delta-9-tetrahydrocannabinol (THC), the psychoactive ingredient in marijuana, the endogenous cannabinoid anandamide, methamphetamine, cocaine, nicotine and heroin. In previous studies we found that caffeine potentiates the motor-activating, discriminative stimulus and reinforcing effects of other psychostimulants, like amphetamine, cocaine and nicotine (Gasior et al., 2000, 2002). We have been interested in elucidating the mechanism of action responsible for this property of caffeine, which could have implications for drug addiction, by its potential ability to potentiate psychostimulant intake. Caffeine is a non-selective adenosine antagonist in vitro, which binds to adenosine A1 and A2A receptors. It is currently believed that blockade of A2A receptors is the main mechanism responsible for its motor activating effects. However, we have recently found experimental evidence for a predominant role of A1 receptors after its acute administration. Thus, caffeine preferentially counteracts the motor depressant effects of A1 receptor agonists at behaviorally relevant doses (Karcz-Kubicha et al., 2003). Furthermore, we have found a preferential involvement of A1 receptors in the discriminative-stimulus effects of a motor-activating dose of caffeine (Solinas et al., in press). Finally, in in vivo microdialysis experiments, the systemic or local administration of caffeine in the shell of the nucleus accumbens produces glutamate and dopamine release. These effects were reproduced by A1 but not by A2A receptor antagonists (Quarta et al., 2004a, 2004b). The experiments with local administration also allowed us to demonstrate that the adenosine-mediated modulation of dopamine release in the nucleus accumbens depends on glutamate neurotransmission and NMDA receptor stimulation (Quarta et al., 2004b). A different mechanism of action for caffeine was found after chronic oral caffeine exposure, which is associated with tolerance to the A1 receptor-mediated effects. Thus, after chronic exposure to caffeine in the drinking water an acute systemic dose of caffeine produced much milder motor activation, which depended on A2A receptor blockade (Karcz-Kubicha et al., 2003). These results were paralleled by results we obtained with in vivo microdialysis experiments. Acute administration of caffeine after chronic oral caffeine exposure did not increase, but instead decreased the extracellular concentrations of dopamine and glutamate in the shell of the nucleus accumbens, an effect mimicked by an A2A receptor antagonist (Quarta et al., 2004a). We have preliminary experimental evidence that suggests that these presynaptic actions on adenosine A1 and A2A receptors are mostly related to the localization of A1-A2A heteromeric receptor complexes in the striatal glutamatergic terminals (Ciruela et al., in preparation). Also at a molecular level, we have found A2A-D2 receptor heteromerization, a postsynaptic target of caffeine localized in the GABAergic enkephalinergic neurons. This involves an epitope-epitope electrostatic interaction (Ciruela et al., 2004). In another series of experiments designed to assess the contribution of endogenous opioid systems in the abuse related behavioral effects of THC we have conducted parallel in vivo microdialysis and drug discrimination investigations. Using in-vivo microdialysis techniques in freely moving rats, we found that THC produces large increases in extracellular levels of a-endorphin in the ventral tegmental area (VTA) and lesser increases in the shell of the nucleus accumbens. We then used a two-lever choice THC-discrimination procedure to investigate whether THC-induced changes in endogenous levels of a-endorphin regulate the discriminative effects of THC. When the opioid agonist morphine was substituted for THC, it did not produce THC-like discriminative effects, but it potentiated the discriminative effects of THC, shifting the THC dose-response curve to the left. The opioid antagonist naloxone reduced the discriminated effects of THC, confirming a facilitatory role for endogenous opioid systems in the discriminative effects of THC. Bilateral microinjections of a-endorphin directly into the VTA, but not into the shell of the nucleus accumbens, markedly potentiated the discriminative effects of ineffective threshold doses of THC, but had no effect when given alone. This potentiation was blocked by naloxone. Altogether these results indicate that psychotropic effects of THC related to drug abuse liability are regulated by THC-induced elevations in extracellular a-endorphin levels in brain areas involved in reward and reinforcement processes (Solinas et al. 2004). These findings suggest a novel mechanism that may underlie many previous observations of opioid system modulation of cannabinoid effects.